Wireless Monitoring Self-Propelled Vehicle Based on Improved Agglomerative Hierarchical Algorithm and Color Space in Image Tracking
| Autor: | Zheng-Hua Yang, 楊政樺 |
|---|---|
| Rok vydání: | 2017 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 105 This paper proposes an improved tracking algorithm by using “agglomerative hierarchical algorithm ” and “ color space ” for real-time tracking of the targets, which combines three methods to realize an automatic tracking system in real-time on wireless monitoring self-propelled vehicles. There are three methods including the background subtraction method, the seeded region growing method, and the wireless monitoring system to construct the color histogram models for each targets. Finally, the obtained image tracking pictures can be saved into flash memory to speed up loading. Due to the advance in computer technologies and the semiconductor technologies, the speed of the processors have been dramatically improved in recent years. More and more complex algorithms then can be used to detect or track objects, and the relating studies flourishes presently bringing about high efficiencies of identifying and tracking objects. It truly makes a major breakthrough in algorithms, and the real-time object tracking system can be achieved. Basically, the object tracking methods proposed in the past years can be generally divided into these categories: region-based tracking, active contour-based tracking, feature-based tracking and model-based tracking. Although visual tracking has been developed for a long time, it still has many problems which needed to be solved. For example, the complex background and the shadow may have an influence on tracking object. This paper uses improved agglomerative hierarchical algorithm and color space for object tracking that can effectively solve the shadowing problems. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|